contribution back to human technological progress. all the stuff people did before. this iphone that i still marvel at every day, all the work that had to go into that. and those people, i'm very grateful to them. they knew they would never get recognition but they also wanted to do something to contribute and so do i. i can imagine better compensation than that feeling. it may be weird if i had already made a bunch of money and planned to make way more from other investments. i just don't think about it. emily: -- sam: i think this will be like -- it will be the most important step yet humanity has to get through with technology. i really care about that.

emily: what is one question you really wish -- sam: people like elon and equity and the personal drama of the day. emily: seriously. i understand you are getting a lot of those questions. sam: i'm happy to talk about it. i'm always excited to just talk about what can happen in the coming few years, few decades with the technology. emily: what are we all going to do when we have nothing to do? sam: talk about elon and zuckerberg i guess. i do think we ever run out of things to do. it is deeply in our nature to want to create. want to be useful. to want to feel the fulfillment of doing something that matters. if you talk to people from thousands of years ago, hundreds

of years ago even, the work we do now would have seemed unimaginable at best and probably trivial. this is not directly necessary for our survival in the sense of creating food or whatever. the shift happens with every technological revolution. we worry about what people are going to do the other side. every time we find things. i suspect this will not be an exception to that. the things we find will be better and more interesting and more impactful than ever before. there are a lot of people that talk about ai like it is the last technological revolution. i suspect from the other side it will look like the first. the other stuff will be so small in comparison. i think the whole thing about technological revolutions is dumb. it is just one long continuous one. it is a continuing exponential. what will be enabled is stuff we

cannot imagine on the other side. we will have way too much to do. if you want to sit around and do nothing, that's fine. emily: bons bons and beaches in my future. we talk about ai designing other ai on the tour. katie play that out for us and the applications? sam: this is the classic sci-fi idea. at some point these systems can help progress themselves. can discover better architectures. can help write their own code. i think we are a ways away from that. it is worth paying attention to. emily: china and russia. you did not go there. sam: i gave a keynote at the chinese conference. emily: you did virtually. where are they are on ai and should we be worried? sam: i don't know. i would love to have better information on that. i don't have a great sense. emily: does it concerning -- it

concern you that we don't? sam: imperfect information is always concerning. i would love to have a better sense. i'm optimistic we can find some sort of collaborative thing. i think this thing in the u.s., which it is impossible to cooperate with china, it's off the table, it is asserted as fact and people are trying to will it into existence but is not clear to me it is true. i suspect it is not. emily: right, right. i'm so grateful you have been around the world talking about this and you are with us today. even you would acknowledge you have an incredible amount of power at this moment in time. why should we trust you? sam: you shouldn't. like, you know, you have known me for a long time. i would rather be in the office working. i think at this moment in time people deserve as much time to

ask questions as they want and i'm trying to show up and do it. more than that, no one person should be trusted here. i don't have super voting shares. i don't want them. the board can fire me. that's important. the board over time needs to get democratized to all of humanity. there are many ways they could be incremented. the reason for our structure and it's so weird and the consequence is we have no equity and we think this technology, the benefits, the access, the governance belongs to humanity as a whole. if this really works it is quite a powerful technology. you should not trust one custody -- company and certainly not one person with it. emily: should we trust openai? sam: only if openai is doing the sorts of things. if years down the road we have

not figured out how to start democratizing control, you shouldn't. but if this is likely figure out some sort of new structure where openai is governed by humanity, and that can happen in many ways, including the alignment data set rather than us picking it. picking it from humanity as a whole. there are many ways. we are talking about what they could look like. emily: thank you for explaining why we should maybe consider trusting you. you have a plane to catch. we are grateful for your time. thank you so much. sam: for sure. [applause] caroline: that was emily chang and sam altman. i'm caroline hyde in new york. ed: i'm ed ludlow in san

francisco. this is a special edition of "bloomberg technology." coming to you live from the bloomberg technology summit. what a way to start. the man of the moment, sam allman. the top level saying we need global regulation. things will go wrong but the future is bright from a technology perspective. caroline: optimistic. some clarity around his own motivating forces. wanting to make impact. wanting to further technology for good and sort of seemingly at odds with what humanity -- why most people want to know why he has not much skin in the game financially. he says he will make money in the future, don't worry about it. interesting perspective on china and russia and the fact that he does not have great insight there at the moment either. ed: me too. the theme of the summit is the

turning point. sam talked about how little we know about what is going on in china in the field of artificial intelligence. he framed it is the biggest step for mankind in terms of technology. i think that will be a discussion here all day long. caroline: we have more ai discussion coming up. the bloomberg technology summit is going on. emily sitting down with reid hoffman and now ceo of inflection ai. let's listen in. >> the digital influences. our belief is everyone will have a personal ai. one aligned to your interests, on your team, in your corner, gets to know you and forms a trusted relationship with you over time. it will be like a confident, a conciliatory -- consigliari. a chief of staff that is your enticing a booking. it will be your digital representative.

negotiating on your behalf, interacting with other sales ais that are trying to encourage you to buy something and helping you to get a great deal. >> for example, is the whole set of things. one thing i found out yesterday is one of our team members at greatbatch is using pi for -- greylock is using pi for parenting advice. >> isn't closer to the customer and a better position to offer that personal digital assistant. >> obviously, i'm on the board of microsoft. i was on the board of openai. part of the thing is what i love about startups is the unique vision where you are unfettered by the other aspects of your business. part of the idea mustapha and the inflection team came up with and i have a little bit of the fingerprint here is iq is not

the lie that matters. matters as well -- eq matters as well. how can this be helpful to you? very high-end iq, but alsoeq. --e q. -- eq. it is how you connect with people. i think that is one of the things the inflection folks are doing better than everyone else. emily: can we talk about the eq thing? i asked pi what we should ask you. the questions were fine. are you going to tell them have this conversation? and it responded haha, i'm just a computer program. i assure you our conversations are confidential. i think it is cool we are having this meta conversation about an interview with my creators. it was the personality that shocked me. can you explain how you designed

that? mustafa: we designed it to be patient, curious, kind. one of the things that first struck us a year and a half ago when we started working on this was what makes for great conversation. when you feel that sense of flow. that sense of energy in connection with another person. i think was to the time it is when you feel heard and understood, he received a little affirmation. it is not completely sycophantic. it does not just agree with you at every moment. he can be a little challenging. it has boundaries. you can push it on certain topics and it will take a position. that is healthy. it is just curious. i think many people are excited by the idea, especially the users we have a someone asking lots of questions about the topic they are interested in. we don't always have someone in our lives who is as knowledgeable and passionate

about our favorite topics as we might like. that is where pi comes in because it is super knowledgeable and engaging. emily: it sounds like a competitor to chatgpt. is it a competitor to chatgpt? reid: i think there will be a number of agents, not one agent to rule them all. like the ring and mordor. emily: will chatgpt will most of us? reid: it is like the same reason we talk to different people for different aspects of our lives. this person we talk about our passion, about snowboarding. this person about what's going on in the country. we have a pantheon. we will have a pantheon of different kinds of ai. one will ask -- you asked chatgpt how you comfort her friend who has lost a treasured pet. here are five possible ways you might do that. you ask pi.

that sounds like it's really hard for you and your friend. what would count as being present for your friend? it has the iq and the five reasons and goes through that style of going through it. these are different interaction experiences. the question is which one do you want this minute, which one do you want more and your life? part of the design of pie is how to help you be the best you. that does not mean being sycophantic. is a question of asking questions. what do you think about this? how do you go through it? help you navigate your life by being helpful in that way. >> talking about the theoretical dangers of each ai and chatgpt 9. i'm curious do you guys think generally that's a discussion worth having and where you are on the spectrum of existential

dangers from ai. reid: one of the things i think is dangerous about existential discussion is that it blinds us from some of the near things. ai's amplification intelligence amplifies human beings. we see a bunch of good things for human beings. ai tutors and doctors. there's also bad human beings. what to do about bad human beings with ai is also part of the portfolio mix of how you do this technology. one of the reasons your book "the coming way" -- i will have the answer ever to you. brad: sam has probably left the building by now. do you disagree with some of the alarmism notes he sounded? mustafa: it is easy to speculate

around what a gpt 9 might look like. six more magnitudes of order of compute would be altering and i'm with those concerns about existential risk. that is many years and unclear what actually happens at that scale. what is much more clear in what explore the book is in the near term we are about to empower many people, if not ultimately everybody with access to the ability to amplify their existing power. not just a knowledge engine but in time it will allow you to take actions. you will make recommendations, buy things, but things. that will be smaller and cheaper and proliferate far and wide. that will cause a dramatic instability and potentially a threat to the nationstate. anyone who has an agenda or is trying to push a political outcome is suddenly going to see the barrier to entry to that scale and the impact

lowered. they will be a question around how states manage that distribution of power. i think there will be a tendency to lunge towards slightly more authoritarian or surveillance-based mechanisms to prevent proliferation, which is better innovation and obviously dystopian from a political government perspective. emily: sam and i were talking backstage. he is much more confident that ai will lead us to a more equal world than an unequal world. are you saying there's a better chance of economic dislocation and social dislocation as a result of this technology? mustafa: both are likely to be true which is surprising. if you look over the last 50 years, the transistor, the last wave that enabled the personal computer has clearly made us in many ways more equal. whether you are a billionaire or $20,000 a year, we all get access to the same cutting-edge hardware. the smartphone, laptop are

broadly at the cutting edge. everyone else will catch up over the next five to 10 years. we are on the same trajectory with respect to access to intelligence. that's an unbelievable idea. over the next decade hundreds of millions of people and then billions of people will get access to the same expert dr., the same expert educator, the same tool for scheduling, prioritizing, organizing your life. that will be the most meritocratic moment in the history of our species for sure. it is a question around how individuals and groups and organizations use that power. clearly we all have conflicting agendas and priorities and goals. that is basically where i think we end up with significant disruption. brad: that is the glass half-full. what is the potential threat to the millions of coders and customer service representatives, the people who are serving jobs whose functions

can be easily replaced by ai? emily: linkedin too. reid: the closest metaphor i have used to describe this moment is it is the steam engine of the mind. if you look at the steam engine, obviously majorly amplified, industrial revolution us superpower muscles. superpower transport, construction, and now we have superpower of the mind. there's a lotta very positive things that come out of that. all the things we have of the increase in wealth that allows medicine, general education, everything else coming out of the industrial revolution, the same will come out of the steam engine of the mind. but the transition is going to be difficult. the transition is going to be, ok, customer service jobs will change a lot. engineers. if you go through a company and go, we tendx beat -- 10x each

function. 10x marketing people. maybe the marketing function will change. let's order entry. you still have a business competitive ecosystem. you don't want to stop marketing or not be doing it. in customer service you get more replacement. here is part of the reason i did the book and i'm trying to orient people. customer service people, there will be transition. ai can help with that. you can build ai to figure out other jobs i can help you find them, help you learn them, help you do them. that is the thing we as a society need to be doing. less how do we slow down ai. how do we shape it to help a broad swath of humanity in this transition? that is where i'm trying to get the dialogue and discussion to. emily: we were just talking to sam about google and whether or not it is still scary. yes, they are still formidable. you quit deep mind at google ai.

obviously, a lot of people have left google. is there something wrong at google or some innovators dilemma will help prevent it from truly succeeding on generative ai because it undermines its own business model? brad: fundamentally, ai stands in tension with google's existing business model. it is hard to eat yourself from within and adapt and respond to the coming wave. for a while google was just idling. i was there. we had chatgpt before chatgpt. it was a remarkable feeling internally, playing with this incredible tool and not really being able to get it out into products. the launch of chatgpt threatened google and shook every thing up. google is a very formidable organization for supersmart people. i am sure they will be just fine but i understand why their

stock. nobody wants -- they are stuck. nobody wants an ai funded by ads. you don't want to salesman in your pocket trying to persuade you to go buy more or do x, y, z. you need fiduciary alignment. your ai has to be on your team and that means ultimately you have to pay for it. if somebody else is paying for it, you have to ask yourself what are they trying to persuade you of? are they trying to influence you in some way? people are alive to that now. that is a real challenge for google. it is unclear how they will manage this transition. emily: we asked sam if google scares him. does openai scare you? brad: no, it -- brad: kid definitely doesn't -- mustafa: we lost our language model, and we set up the bill the model that was fast enough and more capable than every other model on the market for

our computer class. we are very proud i on from our launch we are not better than llama, chatgpt, chinchilla, all the other models of our size. that's important because it empowers pi and ultimately will be available for a conversational api. that demonstrates with a team of 35 people in short order we are able to exceed the cutting edge and now build an absolute best in class ai which is exciting. we have managed to do that because we have been able to gather a huge amount of capital and great investors and train -- we have the largest operational cluster in the world of h100's nvidia,'s leadership. reid: it's an awkward question when you ask sam. should we trust you?

yes, trust me. that's weird. emily: no, you shouldn't. i was like, what? reid: the openai people are great people. there's a 501(c)(3). there's a bunch of social media stuff that obscures that. i spent a number of years on the board in team. i work with them and help them. they are fully paying attention to every serious question. sam spent a month doing a world tour talking to people to say i'm here, you can talk to me about your concerns. i want to make sure it is good for humanity. i don't care about this u.s.-san francisco taxing. -- tech scene. i care about this. it's the kind of reason, in addition to the 501(c)(3). i'm delighted.

brad: i don't see how he's getting any work done. i want to change gears quickly. we will throw some twitter polls on the screen. we asked users where the greatest advancements in ai would come from. i want to ask about the u.s.-china policy of divestment. can be reasonably hope doing that will maybe prolong america's advantage in ai? is a smart strategy? reid: i don't think divestment is a smart strategy. staying connected is better for both the u.s. and china in the world. -- and the world. competition is very good. the fact that there is various ways we in the u.s. have an ai lead. that's a great thing for the values and ecosystem and the great world order the u.s. should be very proud of in the last 70 years. there a lots to be

critical of, of course, but as a peaceful time trying to make connections, and i'm a strong believer in that. investment -- divestment is not the answer. brad: you write china has in a k-swiss at national strategy to be the leader by 2030. do you agree that divestment is a smart strategy? mustafa: china is publishing significantly more papers on ai than we are and the rest of the world. i think our controls were effectively a declaration of economic war on china. they were very firm and aggressive. i think it sets us out on a hyper adversarial footing. they have a huge number of levers and we should expect them to use that against us soon, unfortunately. emily: we talked about regulation. president biden was here this week talking about ai. he met with the presidents,

sundar pichai and satia. what's the real government oversight that protects users, protects the economy, and our political process? reid: the good news is, the administration has taken a little and had to be smart about this. let's assemble people from the industry. yesterday the assembly people from outside the industry. academia and other kinds of places. i think they have a number of good people test with this -- tasked with this. i'm optimistic. on the other hand, a little like your interview with sam, regulatory things is something that is easy to get wrong. you need to be cautious about how you do it. you want to have a net positive impact. not reg to or other things. you have to be careful how you do it.

mustafa: at this time last year almost no world leaders were talking about ai. i don't think there's ever been a technology trajectory in history that is gone from zero recognition to almost universal recognition. that's a good thing. it is in part because we have been trying to advocate and say this is serious and we should pay attention and invite the conversation where it ends up with respect to regulating existential risk or more near-term threats. brad: i think we have to leave it there. thank you guys for joining us. reid: great to see you. ed: reid hoffman, mustafa, the founder of inflection ai the uber technology summit in san francisco. caroline, so much the recap. what they had in common with the idea ai is a technology that typifies or reflects the user.

in society there are good people and bad people. they were not tempted to give the doomsday scenario that we had earlier. caroline: and the eq versus iq. that's a differentiating factor for pi at inflection. they don't feel there is just one chatbot to rule us all. chatgpt will not be the only use case. we discussed that. it feels like chatgpt sucked up all the oxygen. we know there are more focused specific ai chatbots being developed. some with perhaps more emotion than others. ed: the stuff describing ai develop as meritocracy was interesting. let's get back to the bloomberg technology summit where brad stone is on stage with the amazon web services ceo adamson lipsky --adam selipsky. adam: we are by far the most broadly adopted in the world

with the broadest set of capabilities. generative ai is both incredibly explosive and transformative set of technologies in and of itself, and it is fully depended on the cloud to be successful. if you look at the massive amount of compute required -- never mind any other i.t. related stuff. just a massive amount of compute required, that will happen really predominantly in the cloud. companies will want to view generative ai as part of an entire data strategy and data platform. you will want to do your generative ai where you have your data. you also will want the same bulletproof enterprise security privacy you respect from any cloud service. because 70 customers, more than any other cloud, have that data on aws and use it for security and operational excellence. i think they are going to justify the demand that aws have

a full, powerful suite of generative ai services. brad: but do you need a popular or exclusive partner running on aws as it seems like microsoft and google have? who is your horse in the race right now? adam: with all due respect, that's the wrong question. it's like in 1997 when the internet is happening in every thing is going nuts, you and i sang who is the internet company going to be? it seems like a silly question. the leading search company then was altavista. i guarantee my kids are never heard of altavista. ok. who was ahead? which runner is ahead after three steps? it's a 10k race. what everybody needs now is experimentation. venue choice. they need to marketization of generative ai. just like aws was founded to democratize i.t., we need to

democratize generative ai. we have had our own custom chip program for a decade now. way longer than anybody else. not one but two families of chips custom designed for machine learning. those are for people who build models. train models. most customers will direct with amazon bedrock, a managed service for accessing, deploying, managing models. here's where the choice comes in. we are going to have our own models. amazon models lom's have been running inside amazon for a long time now. parts of our retail search are powered by lom's. a lot of alexa's voice responses are empowered. we are in the process of taking this lom's and making them bigger and externalizing them and later this year those will

be exposed for everyone to use. brad: let me ask you -- adam: that is just one set of models. we are exposing anthropic inside a bedrock. stability ai. ai 21 and others over time. nobody knows -- anybody who knows which model is going to be the winner is asking the wrong question. people need to experiment everyone to provide that choice. brad: pretending were talking to a customer now. why should they use titan or one of these other lom's you are exposing rather than chatgpt 4 which seems to have a significant leadership position? adam: i don't know which model they should use. it intends to they are and what their application is. some will probably want to use gpt and some will want to use titan and others anthropic. it is preposterous to think that one model or one company is

going to be the solution for every application and every company out there. we are already seeing this heterogeneity. we are seeing the explosion of interest in bedrock and amazon's generative ai capabilities. this morning the largest advertisers announced they were working with eight yes using bedrock and our custom chips inside of our compute capacity to do generative ai going forward. bbva, one of the largest financial services announced they were working with amazon on generative ai with bedrock. it is that choice combined with the enterprise security and privacy which i think are so fundamental. brad: i can humbly admit as a journalist covering tech for way too long that in the fall i was surprised by not only the quality of chatgpt but the customer response to this new wave of technology. were you surprised? adam: i think -- folks working

in this area of ai have known about largely which models for a long time. very few companies have more experience with ai than amazon. 1998, personalization on the website, that was ai. sage maker, 2017. used by over 100,008 of his customers. most machine learning in the cloud happens on sage maker. we have a lot of experience in people working on lom's. i think the world was surprised with 3.5 came out it was such a dramatic improvement in response. a dramatic improvement in responses over 3.0. that was the surprise but not the overall arc. the good news for our customers is that with deep expertise in ai, working on different forms for a long time, and are now pouring enormous resources into

the generative part of ai. brad: i will give this up after. would you can see that amazon is playing a little bit from behind right now? adam: i don't think so. it is the race analysis. arbery really going to have a conversation about three steps into a 10k race? it is about the long-term. amazon has always taken a much more long-term view of the world in almost any other company. i think the key is building a multiple layers of the stack because that is what customers need. we are building applications on top of these models. we released code whisper, a coding companion on top of amazon lom's. you type in the words and it gives you back code. internal tests coding challenges. developers finished their tasks 57% faster than those not using code whisper. plus, it is secure, private. tells you what open-source you are using and what licensing restrictions there might be on it, was not not every other solution does.

i think we are confident. the thing that ought to give us confidence is customer response. i mentioned omnicom and bbva. really exciting developments to come. we are talking to one company who has millions of lines of mainframe code. they are talking about moving over these really gnarly mainframe applications and millions of lines of code using generative ai from aws. it is this consumer application, this chat application that is easy for folks understand because they can say give me a haiku about farm machinery. it doesn't. brad: that is the key, the haiku's. adam: what folks in this room and watching online understand is there is a full suite of

enterprise and company and organizational applications. they will be huge needs there. we will be focused on customer service and coding. and drug discovery. a full suite of applications across every industry. i think we are well-positioned. brad: do you feel like you are close to having a gpt for quality model running on aws? adam: all the tests we have done and the customers in the existing, current preview of bedrock have been very impressed with the quality of our models. it is not just about our models. we will be proud of our models. anthropic doesn't amazing job. they are right up there -- does an amazing job. stability ai for models generating images. collectively, these models

provide the best destination. for consistent api set, consistent data be a security, consistent identity systems on a private, isolated, virtual private cloud. none of this application and now you have to have a bunch of fortune 500 cios ban it from their companies, which has happened. from day one it's always going to be aws class security. brad: i want to put a slide of showing bloomberg intelligence estimate on projected generative ai revenue in the next maybe seven or eight years. who knows about the numbers. it is up to the right. i can imagine a lot of financial types who are out there looking at amazon's numbers and wondering how much of a sales tailwind this will be for a ws. what can you tell us about -- you mention it competition --

computationally for aws. adam: prognostication is fun but it's also important to be humble and nimble. anything ending in -le. we have to adjust rapidly. i think that -- look, i don't any fundamentals about cloud computing have changed. who knows. call it 10% of i.t. has moved to the cloud. we are very early. whether you are talking about any application there is massive runway for things to move to the cloud. we firmly believe they will. on top of that, i think that generative ai will be the next massive increase in workloads moving to the cloud, or in many cases happening for the first time and happening in the cloud. i do think it should be a significant tailwind for cloud providers and particularly for

aws given our leadership position. we need to come up with the capabilities, the services that justify people using us for that purpose. i think if we do a good job and listen to customers it should provide significant demand for years to come. the computational requirements are so intense. one thing that works in favor of aws is that a lot of people ask about the energy consumption. what about sustainability? brad: you run those efforts insight -- inside amazon. adam: we as a company and i care a lot about sustainability. people say, well, generative ai, massive compute, is that incompatible? no, we are not. these workloads, there is no putting the genie back in the bottle. generative ai is going to happen so let's make it happen in a highly energy efficient and sustainable way. if you look at our custom chip

we design, if you take rabbits on, our oldest g -- graviton, it is 60% more energy-efficient. if you look at an enterprise in general, never mind generative ai, but moving from data centers to the cloud, a study showed aws's 3.6 time more energy-efficient than the average enterprise data center in the united states. we are the sustainable place. we are doing it through a series of technological improvements, plus a commitment we are 85% of the way to being 100% renewable energy by 2025, which is just around the corner. brad: you had another commitment called chip in zero, and that carbon zero commitment by 2030 which you guys maybe scrapped or took off the website. i was a longtime amazon watcher

and my heart sank. you guys have been so prominent about these goals. the question is, have these commitments become harder to meet than they were to make? adam: your heart should be thinking at the state of global warming. brad: it certainly is. adam: we should all be concerned about that. your heart should be singing at the leadership position amazon is trying to establish and the improvements we are trying to make in the public goals we have set. brad: why did you remove that commitment? adam: it is subsumed in an older commitment. he made a public place to be net zero carbon across all of amazon, not just aws, by 2040, which is 10 years ahead of the paris accords. for a technology company like aws, i will not say it's easy but it is doable. you will hear that from other tech companies. for big retailers with airfreight and inbound transportation and stores and packaging, it's really hard.

we will be of the first to say we don't know how we are going to get there in all dimensions. i can tell you about renewable energy, not about elements of transportation and packaging and buildings. we are an innovative company. we take bold long-term bets having made this pledge publicly, not privately, to number one have a force in function for ourselves because it is not easy. also, we want to inspire other organizations, governments, companies to join us. we have 400 signatories with the climate pledge that is growing every month. it is not a competition against other organizations. it's a competition against the thermometer. i want people to out innovate us. i want to be beaten in the race to become sustainable. hopefully we can be inspired by things other people are doing. we adjust our goals over time but the thing i focus on is the audacious goal to be net zero carbon by 2040. brad: and the line in the sand.

things like punitive ai take more computational resources. even as we get more efficient with compute, that's a commitment you can hold firm on, even if you don't know how to get there? adam: we intend on getting there. there are lots of other areas of progress we are making in the interim. if you take packaging for the retail business, the average packaging per shipment has decreased by 38% since 2015. which is another example that it can be sustainable and good for business. lower cost for us and much more sustainable for the planet. anytime you create a win-win like that it works for everybody. it becomes a sustainable business proposition. brad: let's get back to ai and alexa. it is slightly out of your purview but alexa runs on your servers. can alexa be a generative ai place? should it be? adam: alexa is already powered in large part by lom's amazon

built and has been in production for a while. i think she's only going to get smarter and better and more personalized. brad: it feels to me as a longtime user knows i have populated our house at one point but today it feels as though it's a step behind. would you agree with that? adam: i think -- i use alexa in my house. different strokes for different folks and we love alexa. i like to think she loves us. she will not actually tell you yes. alexa has been getting better and better. more skills, better skills. more understanding of you. the rapid improvements making a generative ai are truly going to continue to transform alexa into

a truly personalized assistant. we want alexa to be an absolutely indispensable best in the world personal assistant to you. we have a lot of work to do. we want alexa to fill in the blends over time but we are confident in that plan. very optimistic about alexa being able to fulfill that role in people's lives that they are really going to love. adam: i want to get to two more. the processor and cutting some of the big bets at amazon. one bet he has not cut is the satellite plan, kuiper. what dcs the opportunity considering rival systems star link from tesla. it seems to be operational in quite far ahead. adam: we are optimistic about kuiper. there is huge interest from government and organizations. billions of people around the world are underserved for

internet. kuiper aims to democratize that. to democratize that and provide great internet service to somebody billions of underserved people. in addition, whether it is automobile companies, telecommunications companies, but of other enterprises for a w's customers and government -- for aws customers and governments to be able to backhaul information up to kuiper, back to the aws cloud. as we launch our first satellites later this year and really ramping up in 2024 and 2025 as is my understanding, but then being in-service in that timeframe and being able to deliver that for aws customers is huge. i do just want to remind folks that just this morning aws launched its $100 million genitive ai innovation center. we will be going out to all

those customers around the world, enterprises, with expertise, free aws solutions. architect, engineers, strategists, and working with them one-on-one to envision, design and -- genitive ai capabilities. not talk but actual -- brad: discount for new companies? adam: we will bring our internal aws experts free of charge to a bunch of aws customers. folks with significant aws presence. it will help them turbocharge their efforts to get real with genitive ai. get beyond the talk. brad: last one in -10 seconds. you and andy are in the same situation. you are the guy after the guy. the founder whose name was synonymous with the early-stage of legendary growth. what do you want adam's aws legacy to be? adam: i don't really think of it in personal terms.

i don't have a can answer. -- canned answer. i would love it if i could be known to really help drive business that is constantly, no matter how big he gets, how far-flung it gets, puts customers at the center of what we are doing. always put customers' interest before anybody else's interests. it's an empathetic, equitable, fun and innovative place for employees to work. adam: adam selipsky, thank you for joining us. caroline: adam selipsky in conversation with brad stone. highlighting some of the news that has just come out. the fact that amazon is spending $100 million to teach cloud clients about ai, to get real basically in the early-stage clients. they will be really using some of these customized applications and understanding how to get this expertise to ensure they are adopting genitive ai at a

rapid rate. we heard about the focus on being carbon neutral. focused on climate and how you twin that with the enormous compute power necessary as we dive into the world of language models, more generative ai, and chips necessary, the computer power and what that efforts -- how that affects the climate. the conversation just keeps going. liver technology summit with ed ludlow sink sitting down with cristiano amon. ed: taking photos. the question is qualcomm and ai company? cristiano: catch a great question to ask. it's incredible to see all the development you see now on ai. here is how i will answer that question. equity is very simple. if you think about the ai,

semiconductors accelerating computers, a lot of competition. what we see and what you can do with those line which models, large models for images and video. if you think about the history of computing, computing starts in the cloud and get scaled at the edge. that is what happens with cpu's. that is what happens with all other form of computing. the smartphone is a great example of that. the largest computing platform ever developed is the smartphone right now. the largest develop platform for mankind. what is good about the smartphone, it's a device that is with you all the time. if ai becomes pervasive, which we believe it will become pervasive, especially you look at how those large models are very natural. you can converse with them. they have contextual information

and all those things. that will happen at the edge. that is how you should think about qualcomm. if ai is going to get scaled, you will see it running on qualcomm snapdragon devices. whether it is on your phone, your car, your pc, other machines. it's a great opportunity for us. ed: the future is to democratize access to artificial intelligence and those generative ai tools. qualcomm will make that happen. why are you not getting like -- cristiano: i think what is happening now -- by the way, it is great for the semiconductor industry. anybody that has been on the forefront of computing. qualcomm is well-known as a communication company. if you look at what were doing now it is more of a connected processor company and communication. as those models started to become very popular, they will be running at the edge.

i expect that ai becomes an option for qualcomm right now. i will give you an example. i saw something adam said in the prior conversation when he said something about in 1997, who could guess who were the winners and losers on the internet it would be a wild guess. genitive ai opportunity is huge. we don't know yet all the different applications that are going to come up. just within the past six months it's a revolution in the number of companies with use cases. those are going to happen on devices. i think that is going to be a great opportunity. ed: i'm going to show you something to the audience here and those with us virtually. think about questions for cristiano based on what you see. let's bring up the video. when it comes up and plays,

explained it was a little bit what it is we are seeing. here at the bloomberg technology summit we will nail the technology. just wait. the video is going to come. and when it does it will have been worth the wait. irika. -- here we go. cristiano: you have an input image on your phone. you tell the input prompt what you want the image to be. make it a masterpiece, look like the venice canals, 4k. it just runs and he gives you this very unique image. image to image, that has never bankrate a before through ai running on your phone. it's a good time to talk a little bit about how we think about ai at the edge, outside the data center. like we have seen everywhere there will be this huge opportunity for the cloud but it would be this huge opportunity for devices. the device is very different.

there are a number of reasons why this will be very popular on the device. first, the device has contextual information about you. it has real-time information. like the picture you just took. you want it right now at that moment, change the picture and share with some of the else with your messaging platform. ed: that video, that device was run in airplane mode without any external connection. it ran the model locally on -- cristiano: absolutely. real-time contextual information. there's another reason. processing on the phone is virtually free. when you think about running those models in the cloud and think about large language models, for every token like a word in a sentence, if you have an experiment that you see the words coming up-- caroline: the president and ceo of cristiano amon discussing how

qualcomm will play a role in the enormous scope of generative ai, and computing. the chips that are known to be in your iphones and phones and you're all that mobiles -- automobiles, qualcomm wants to make sure you're accessing genitive ai in local ways and means. an amazing conversation that continues. go to live go if you have a tournament or go online to watch the conversations. so many more from the bloomberg technology summit in san francisco. right now you can stick with bloomberg television. we will dive into the world of bloomberg equality. for now, let's leave it with cristiano amon. ♪

when you automate sales tax with avalara, you don't have to worry about things like changing tax rates or filing returns. avalarahhh ahhh we moved out of the city so our little sophie could appreciate nature. avalarahhh but then he got us t-mobile home internet. i was just trying to improve our signal, so some of the trees had to go. i might've taken it a step too far.

(chainsaw revs) (tree crashes) (chainsaw continues) (daughter screams) let's pretend for a second that you didn't let down your entire family. what would that reality look like? well i guess i would've gotten us xfinity... and we'd have a better view. do you need mulch? what, we have a ton of mulch.

caroline: this is bloomberg equality. i'm caroline hyde. romaine: i'm romaine bostick. we are taking an all-inclusive look at what social equity and equality truly mean for companies, investors and our collective global prosperity. we will go deep into several key topics, including politics around gender identity and fair access to capital. caroline: today, that is why we start with access to capital. the global economic landscape has changed, aftershocks from the pandemic, the war in ukraine, the bank